Everywhere you look these days you are seeing AI being used. From university students using ChatGPT to write essays, to a colleague using it to write his leaving email, it’s here! So whilst it’s a bit off topic from my usual blogs about running, I decided to have a play and then write about AI (important thing to note is that this post is all human written!)
Having read so much about ChatGPT, I had decided to sign up and have a go myself about a couple of months ago. Could it be a useful tool, especially if I wanted to make this blog work better.
It’s worth at this point to give a brief background of me and my online history. For many years I ran a website that helped Gluten Free Travellers. Unfortunately though the costs of hosting the site and the time required to maintain it had become too much so I had to make the decision to close the site in January. Through running this site (and a bit through my day job) I have developed a bit of an interest in SEO. This blog is primarily a conduit for me to share my thoughts but no-one likes screaming into a void, so I am always on the lookout for ways to get a couple of extra reads of my posts.
Therefore my first forays into using the tool were simple top level requests like “how to make my blog more successful”. The output was valid, but no more than I’d have found in a normal Google search and very top level. I did even try and ask it to write a blog piece for me. I’m sure to someone else it would read fine but it didn’t sound like me to me so I won’t be going down that path any day soon.
However, I did still feel it must have the power to help me somehow. A little while later I discovered a new SEO tool called TopicRanker. It uses a clever algorithm to find keywords that are relevant to your site but the current Google results are quite poor. You only get on free search per month but it got me wondering, could ChatGPT help.
ChatGPT – Help Me!
My opening question was as follows: “I’m looking for keywords relevant to my website that get at least 10 searches a month but the current top ranking Web page has a low authority score“. The results were a little generic and in the past this is where I’d given up.
This time though I probed a little more: “Can you help me find some examples“. I should have been more specific as I got some very general results like: “How to train a puppy to sit”. If this was relevant to my blog it would have been great but my question needed refinement.
Next up: “Can you help me with some examples relevant to my site mikeruns.co.uk“. Finally started to get somewhere with the answers!
- “Marathon training plans for beginners”
- “Best running shoes for long distance running”
- “How to prevent running injuries”
- “Running nutrition tips for endurance athletes”
- “Tips for improving your 5K time”
All of these were good suggestions and relevant but at this time it felt that writing a marathon training plan when out of training or nutrition tips may not be as relevant so I tried one more time given that my most visited pages have been parkrun related: “Any examples linked to parkrun“
And there I had it, a curated list of ideas, all of which I thought could be relevant and I could do a piece for.
- “How to prepare for your first parkrun”
- “What to expect at parkrun”
- “Benefits of parkrun for mental health”
- “parkrun warm-up exercises”
- “parkrun tourism guide”
I settled on the last keyword as it seemed something I could pull together quite quickly and if I could rank better for it, perhaps it could help get some more regular, organic traffic.
I did then try to use the AI to write the post for me but as mentioned earlier, I didn’t really feel it sounded like me but, using some of the suggested layout as a guide, I quickly wrote my page to target the key word “parkrun tourism guide“.
Did It Work?
This is all well and good but, did the foray into AI work? To work this out I looked at the stats for my site on Google Search Console. Now my traffic is pretty low by any standards but in the week before I published the AI inspired page, I showed up in Google Search just 34 times. In the 7 days after this rose to 71! Now only 12 of these were the new page itself but then SEO is a complex beast and the halo effect of this page seems to have helped with my ranking for other parkrun terms. What’s more, if I search for the specific term “parkrun tourism guide”, the new page is already ranked #4 (even if I check in incognito mode to rid myself of any bias). For a brand new page that is impressive and shows that the AI results suggesting it as rich pickings was a good idea. However, despite this, I haven’t had any more clicks to the site so maybe it’s too soon to jump for joy!
Does any of this matter?
This entire exercise was really just a chance for me to play with ChatGPT. The real success of a blog will come from regular posting and building a following. A slightly better position in search won’t harm me but given it currently needs about 50 search impressions per site view, its still quite a way until I could drive big volumes. It’s also true that nothing that I did with the AI couldn’t have been done with a bit of my time. However, the fact that I was able to identify a niche keyword, frame a post and then use that frame to write and post within a short time, and that post shows early signs of success makes me think I will look for some AI help again in the future, even if I don’t come to rely on it yet.
If you’ve used AI to help on your blog at all and have any good ideas then please do post them in the comments!